Enhanced case detection of alcoholism is a major component of any efforts designed to prevent the unfortunate consequences of the disease. Many alcoholics are admitted to general hospitals with diagnoses not apparently related to alcohol abuse, and these persons may have no other access to alcoholism care. Thus, the general hospital offers unique opportunities for developing both improved detection methods and the means for their integration into the treatment referral system. The work proposed here aims to apply modern decision theory and multivariate analysis to demonstrate: 1) that detection of alcoholism in hospitalized and ambulatory patients at a general hospital can be markedly improved by properly directed utilization and interpretation of existing multiservice diagnostic information, and 2) that incoporation of this diagnostic information into the process of clinical decision making affords a new approach to remedying the well known educational deficiencies in professional training concerning alcoholism. Success of nationwide efforts to establish effective alcoholism programs in all general hospitals is dependent upon reliable, sensitive and objective methods for identifying the alcoholic patient and for improving the medical decision making processes as they are applied to the patient suffering from alcohol abuse. The proposed research will develop and test a clinical diagnostic algorithm which should facilitate earlier detection of alcoholism and enhance professional appreciation of effective diagnostic maneuvers. BIBLIOGRAPHIC REFERENCES: Jankowski CB, Drum DE: Diagnostic Correlates of discharge against medical advice. Arch Gen Psychiatry 34:153-155, 1977. Jankowski CB, Drum DE: Criteria for the diagnosis of alcoholism occurrence and limitations in general hospital patients, in Seixas FA (ed): Currents in Alcoholism, Vol I. New York, Grune and Stratton, 1977, pp 341-359.